Dear All, I have two questions:
1) I have collected multiple, native datasets (5) from the same crystal (different parts of the crystals exposed with different transmission and oscillation angles). Each dataset on its own is close to complete (96-98 %). Naturally, differences in exposure, onset of radiation damage (datasets were collected with high transmission), local differences in the crystal, will affect the variance of the errors in the measurements for the reflections between the different datasets; but I would think the redundancy and increased number of measurements from all datasets should outweigh this. My tendency is to include all datasets.
I am working at 3.8 A resolution (structure is solved; 80 % solvent content). Missing even a few reflection will probably have more of an impact at this lower than at higher resolution. Essentially, I am trying to obtain better signal in the resolution range 4-3.8 A, where there is also diffuse scattering from the solvent between 4 and 3 A and ice rings and the signal from the individual datasets is weak. Obviously the criterion will be the calculated map quality, but wanted to know what some experiences of people have been in such cases. Should I merge the datasets or rather use them individually for map calculations?
2) What's the quickest/easiest way to ensure equivalent indexing in ccp4/imosflm/scala, when merging different datasets together (space group P6222) (in XDS there is reference_data_set). Use pointless then cad+scala?
Cheers, Florian
